[REXNFX]

Looking forward to some examples of good vs bad.
How would a DAC bother about the quality of a recording, how would it even know what's good or what's bad?

Don't know how DAC's work eh? Check this out from 16mins:

 

[gregorio]

To verify our assessments, we can measure the SHINAD of each car so as to make more objective the process of determining vehicle shininess.

Such is the Head-Fi way...

Can’t really agree with that. “The head-fi way” would most commonly be to not verify our assessments at all! It’s our experience/opinion and is therefore unquestionable (doesn’t require verification).

If there is any verification then it certainly wouldn’t be with any sort of measurement, because although we might be able to measure say light reflectivity, that has little or nothing to do with how we perceive and feel about shininess. The way to verify is more likely to be: Proudly show your wife your shiny, new, more expensive car and see if her opinion matches yours. Additionally, the verification level is higher if she doesn’t know much about cars!

G

 

[wantan]

Don't know how DAC's work eh? Check this out from 16mins:

I'll leave the details how DACs work to the experts, who will eventually comment on the video, too. This is typical audiophile talk, lots of technical chatter without any substance. It's just advertising their own "competence". There are no ultrasonic noise issues in DS DACs, otherwise the industry would have solved these issues already.

It also doesn't answer the question how the "quality" of a recording would relate to the "quality" of a DAC. You claimed it would take a good DAC for bad recordings to sound good. Best you post some examples what you hear, good and bad, and what audio equipment you use to determine that.

 

[paulrbarnard]

I'll leave the details how DACs work to the experts, who will eventually comment on the video, too. This is typical audiophile talk, lots of technical chatter without any substance. It's just advertising their own "competence". There are no ultrasonic noise issues in DS DACs, otherwise the industry would have solved these issues already.

It also doesn't answer the question how the "quality" of a recording would relate to the "quality" of a DAC. You claimed it would take a good DAC for bad recordings to sound good. Best you post some examples what you hear, good and bad, and what audio equipment you use to determine that.

I suspect this could be related to upsampling

 

[gregorio]

Don't know how DAC's work eh? Check this out from 16mins:

That video explains how DAC design evolved in very basic terms, although he doesn’t actually give the time line and it ends in the mid 1990’s.

However, your reference of how DACs work (the video “from 16mins”) doesn’t even mention good or bad recordings, let alone how a “great” DAC would differentiate one from the other and then make the bad ones “more acceptable”.

So according to your own reference, you “Don’t know how DAC’s work eh?”

G

 

[Ryokan]

Yes. I can hear the difference. But usually you have to jump up in price point. A $200 dac is probably going to sound like every other $200 dac. But a $20,000 dac or higher is not going to sound like a $200 dac. Also, a $100,000 dac will not sound like a $20,000 dac.

So a rich person can safely buy the most expensive dac in the shop knowing it's going to sound superior to everything else without even testing it in their system?
That rational goes against everything I've learnt in the science forums.

 

[wantan]

If you don't level match very closely DACs will always sound different, the real question is if they still sound different at the same output level. I've listened to many DACs over the years and comparing without level matching doesn't get me anywhere. For some years I though it would work with long time testing but for me it doesn't.
You need level matching and you need to be able to switch instantly between A and B if you want to hear the truth. Whatever the truth is.

I've used a Schiit Fulla 2 as basis to compare the higher end setups. There should be a very obvious step up between Fulla 2 and a 3k+ DAC/Amp setup with all the bells and whistles but there wasn't. Fulla 2 was always to close for comfort.

 

[BS5711]

To extend this a little. It’s not just about the comparison in raw performance, which can be measured, that plays a part. The other factor to consider is that your ears provide a limit. So it becomes comparing the top speed of vehicle on a road with a 70mph speed limit. Sure the Tesla may be 10 or even 50mph faster than the Prius but you can’t use it above 70 anyway.

When you consider the whole chain and include the amplifier and the headphones you might well then notice a difference. Consider the headphones to be a trailer. Some are big heavy trailers (susvara) some are tiny little light ones (Shure IEM). Now it’s quite possible the Prius will only get to 50 with a big heavy trailer tagged on behind and might even start to make more motor noise as it tries to do it while a Tesla might still make it to the 70 speed limit.

Didn’t you just draw an analogy with amplification not digital to analogue conversion ?

 

[gimmeheadroom]

You claimed it would take a good DAC for bad recordings to sound good. Best you post some examples what you hear, good and bad, and what audio equipment you use to determine that.

It's actually quite the opposite. The better the DAC is, the worse bad recordings sound. And that's how it should be.

 

[paulrbarnard]

Didn’t you just draw an analogy with amplification not digital to analogue conversion ?

For the trailer part yes. The clue was in “the whole chain”

The first part referred to the DAC

 

[gregorio]

It's just advertising their own "competence".

To be fair, most of what he stated (“from 16mins”) is actually true/correct, although rather oversimplified. From around 22mins, it’s more advertising and diverging from the relevant facts.

There are no ultrasonic noise issues in DS DACs, otherwise the industry would have solved these issues already.

There is an ultrasonic noise issue with DS DACs but you’re right that the industry solved these issues already. In essence, the point is to “trade-off” a complex, relatively inaccurate reconstruction filter for a simpler/cheaper more accurate (less artefacts) filter at the expense of adding more dither noise. This additional dither noise was dealt with by moving (shaping) it to the (inaudible) ultrasonic range. The issue with 1bit DS DACs (introduced in the late 1980’s) is that this “noise-shaped dither” cannot be fully applied with just 1bit available and therefore a tiny amount of distortion is possible, although in practice it’s inaudible. The industry solved this (inaudible, potential issue) a few years later in the mid 1990’s by using more bits, typically 5-6 bits. Noise-shaped dither could be fully applied, no more potential issue, end of story.

G

 

[wantan]

To be fair, most of what he stated (“from 16mins”) is actually true/correct, although rather oversimplified. From around 22mins, it’s more advertising and diverging from the relevant facts.

There is an ultrasonic noise issue with DS DACs but you’re right that the industry solved these issues already. In essence, the point is to “trade-off” a complex, relatively inaccurate reconstruction filter for a simpler/cheaper more accurate (less artefacts) filter at the expense of adding more dither noise. This additional dither noise was dealt with by moving (shaping) it to the (inaudible) ultrasonic range. The issue with 1bit DS DACs (introduced in the late 1980’s) is that this “noise-shaped dither” cannot be fully applied with just 1bit available and therefore a tiny amount of distortion is possible, although in practice it’s inaudible. The industry solved this (inaudible, potential issue) a few years later in the mid 1990’s by using more bits, typically 5-6 bits. Noise-shaped dither could be fully applied, no more potential issue, end of story.

G

Thanks for filling out the blanks!

 

[BS5711]

For the trailer part yes. The clue was in “the whole chain”

The first part referred to the DAC

In that case your analogy is utterly lost on me.

Isn’t the maximum speed analogous to to total system performance as in sound quality and the DAC would be the speed controller and the amplifier would be the motors and the volume would be the speed you chose to drive at not the maximum possible speed ?

Therefore the DAC is only part of the chain and as in the speed controller may have no effect on the sound quality/speed potential so long as it operates correctly.

Regardless, it doesn’t matter, I was just curious if I was missing something, please don’t try to explain it again.

Thanks

 

[REXNFX]

That video explains how DAC design evolved in very basic terms, although he doesn’t actually give the time line and it ends in the mid 1990’s.

However, your reference of how DACs work (the video “from 16mins”) doesn’t even mention good or bad recordings, let alone how a “great” DAC would differentiate one from the other and then make the bad ones “more acceptable”.

So according to your own reference, you “Don’t know how DAC’s work eh?”

G

Eelco Grimm is one of the preeminent designers of digital audio in the world, and you are?

 

[paulrbarnard]

In that case your analogy is utterly lost on me.

Isn’t the maximum speed analogous to to total system performance as in sound quality and the DAC would be the speed controller and the amplifier would be the motors and the volume would be the speed you chose to drive at not the maximum possible speed ?

Therefore the DAC is only part of the chain and as in the speed controller may have no effect on the sound quality/speed potential so long as it operates correctly.

Regardless, it doesn’t matter, I was just curious if I was missing something, please don’t try to explain it again.

Thanks

I thought it clear but hey ho. My point was the frequency response, the dynamic range etc of the DAC goes well beyond what can be detected by the ear. So a more expensive DAC could measure way better than a cheap DAC but you can’t hear the difference because the extra ‘performance’ is not usable.